Chapter 9: Building, Training, and Validating a Linear Model
In the last chapter, we learned about statistical data tests to validate whether it is useful to build a predictive model. It depends on the strong relationship between the predictor variable, X, and the result variable, Y.
In this chapter, we will develop a prediction model to see whether the amount of fuel in miles per gallon is affected by motor horsepower. We will understand how to measure the difference between the expected values from the data source and the response given by the model.
The strong or weak relationship between the expected values and the linear regression is based on the difference between the training data and the regression model. We can use several tests to check whether the predictor variable and the result variable have a link. These statistical tests can be as follows:
- The coefficients of determination and correlation
- t-statistics
- The p-value
- f-statistics
t-statistics...
